INDEX

Explanations

negative constructs and restrictions in the text

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

soDeliveryDate

-0.82

Kings

-0.71

quickShipAvailable

-0.68

 Lessons

-0.68

haven

-0.67

Luck

-0.66

reality

-0.66

 realities

-0.65

 Niet

-0.65

Lind

-0.65

POSITIVE LOGITS

 permitted

0.88

 interfere

0.87

 exceed

0.86

 discriminate

0.83

 participate

0.82

use

0.79

 duplicate

0.79

 reproduce

0.77

orem

0.75

 allowed

0.75

Activations Density 0.123%

negative constructs and restrictions in the text

No Comments

No Known Activations